LAUSR

Accurately quantifying the spatial and temporal patterns of regional evapotranspiration and their influencing factors has significant implications for understanding carbon and water cycles in the terrestrial ecosystem. Based on MODIS remote sensing data and meteorological data from 2000 to 2015, we used the modified Priestly⁃Taylor Jet Propulsion Laboratory (PT⁃JPL) model to simulate evapotranspiration and its spatial distribution in the Chinese terrestrial ecosystem and analyzed the temporal and spatial variation of evapotranspiration and its influencing factors. The results revealed the following. 1) The observation data demonstrated that the PT⁃JPL model with optimized parameters illustrated 68% of the seasonal variation and spatial distribution patterns of evapotranspiration better than the original model and the MODIS evapotranspiration product. 2 ) The mean evapotranspiration over the years was 440. 16 mm / a with obvious spatial heterogeneity, gradually decreasing from the southeast coast to the inland of the northwest. 3) The total evapotranspiration showed a slight downward trend (slope = 6.48 Gt / a, P = 0.17) from 2000 to 2015; however, there was an interdecadal difference as the evapotranspiration significantly decreased from 2000 to 2010 (slope = 21.05 Gt / a, P<0.01). The Inner Mongolia region, Loess Plateau region, Gan⁃xin region, and Qinghai⁃Tibet region accounted for 45.05% of the total national evapotranspiration and 61.88% of the interannual change; these regions showed a slightly increasing but insignificant trend from 2010 to 2015 ( slope = 10. 48 Gt / a, P = 0. 71). 4) Radiation was the main influencing factor of the trend of evapotranspiration before and after 2010, accounting for 51.45% and 85.26% of the total, respectively.